Regional Study On Meteorological Drought And Drought Vulnerability

With the intensification of the global drought trend,extreme drought events have occurred frequently in recent years,which has aroused global attention.China is a country with frequent droughts.The severe drought has brought challenges to drought mitigation,and there has been a requirement of shift from crisis management to risk management.Drought vulnerability assessment is the basis of risk management.Therefore,it is urgent to develop a scientific assessment model to determine which results in high vulnerability.Thus,establish a drought management based on disaster prevention and preparedness.Taking Xinjiang Uygur Autonomous Region as an example,this paper analyzes the temporal and spatial change characteristics of meteorological drought,and constructs a multi-attribute assessment model for drought vulnerability.The main research contents and conclusions are as follows:(1)Based on the data of 39 meteorological stations from 1970 to 2018,the trends of precipitation amount and precipitation structure were analyzed by Trend-free Pre-whitening Mann–Kendall test,Sen’s slope and Partial Trend Method(PTM).The results show that the precipitation in Xinjiang has a privailing trend of growth,and mainly occoured in the north of the Tianshan Mountains.But in different region of the north of the Tianshan Mountains,the increasing precipitation is due to the increase of precipitation in different seasons.In contrast,in the south of the Tianshan Mountains,the precipitation in the western and southern Tarim Basin is increasing due to the increase in July.About half of the stations show an increasing trend of rainy days,and mostly occurred in the west and north.In constrast,simple daily intensity index(SDII,average precipitation from wet days)shows a significant trend only in a few stations,which implies the increase in the total precipitation is mainly caused by the increasing rainy days.The overall trend of consecutive dry days is not significant,but the PTM detectes a significant decrease in extreme consecutive dry days,which is hard to find by traditional trend analysis methods.(2)In different regions,a drought may not perform in time consistently.Therefore,dynamic time warping(DTW)is used to scale the time axis,and esitimates the similarity of the actual fluctuations of drought indexes at each station.Based on hierarchical clustering,the drought changes are divided into three types: north,centre and south.This calssification has broken the traditional North and South Xinjiang zoning schemes.Based on the classification,drought record testifies the standardized precipitation index(SPI)at the 6-month scale is a good indicator of the actual drought.The Morlet wavelet transform method is used to study the periodicity and abruptness of the 6-month scale SPI.From high to low oscillation intensity,the main periods in the north are 4.3,12,8.5 and 2 years,the main periods in the centre are 12,16,6,and 5 years,the main periods in the south are 11,4,and 8 years.The abrupt changes obtained by wavelet analysis are consistent with the actual situation.Currently,the north is in a positive phase of fluctuation,and drought may be aggravated in the future;the centre and south are in a negative phase of fluctuation,and drought may be alleviated in the future.Using the station proportion of drought and drought frequency statistics to analyze the extent and frequency of drought,it is found that the seasonal-scale global drought is not common,and the annual-scale global drought often occurs in the north and centre.These indicate drought dose not occure intensively during the year.Seen from the drought frequency,spring drought,summer drought and autumn drought are most likely to occur in the south,north and centre,respectively.The centre and south are most prone to heavy drought.(3)A multi-attribute assessment model for drought vulnerability is established based on the overlay and index method.The min-max normalization is popular for data preprocessing,because it retains the information of the original data.However,it only suits for the linear relationship.Abuse will make the result unreasonable.This paper constructs a nonlinear standardization method with adjustable curvature.This method not only makes full use of original data information,but also is applicable to various relationships between impact factors and drought vulnerability.For the standardization process that enlarges the index score to a uniform range,the power factor is used to correct it.The interquartile range is used to amend the abnormality structure of scores caused by outliers and data skewness.This paper selects 10 indicators including precipitation deficiencies,evapotranspiration,water yield property,river density,soil texture,soil organic matter,slope,land use type,population density and GDP.The normalization of indices is processed by nonlinear standardization method with adjustable curvature.Indices are selected based on the information and correlation of the indices.According to the order relation analysis method and the mean square error method,the subjective and objective weights are combined to establish integrated weights.The weighted arithmetic average operator is used to establish the composite index to represent the drought vulnerability in Xinjiang.The goodness of variance fit and Jenk optimization method are applied to determine the number and scheme of vulnerability classification grades.Then,the drought vulnerability map is produced.Based on sensitivity analysis,the contribution ratios of impact factors to drought vulnerability in different regions are analyzed.The results indicate that high vulnerability is concentrated in unused land.Due to relativity of the vulnerability,vulnerability of the used land has been assessed independently.It has been found that the high vulnerability is located in the mountains of the Kunlun Mountains in Kashgar,Artux,Akqi,Uchturpan and plain in Ruoqiang County.Changes of drought vulnerability from 1995 to 2015 were studied based on the dynamic indices inculding precipitation deficiencies,land use types and population density.The results indicate that drought vulnerability has increased in most parts of Xinjiang.Because of importance of agriculture,forestry and animal husbandry in Xinjiang,grassland,woodland and cultivated land are extracted and studies respectively.For the cultivated land,the oasis belts in the north of Tianshan Mountains,the Aksu Plain and the Hotan are analyzed specifically.Finally,based on the assessment results,drought mitigation strategies are proposed for different regions.